Bug 45334

Please attach your xorg log and dmesg output.

Created attachment 56256 [details]
dmesg

Created attachment 56257 [details]
Xorg.0.log

Done

I can confirm this on a Dell vostro 3555 Notebook with A8-3500M 
VGA-0 will be detected if plugged in by kde control center, but no output signal goes out.

3.2.1-gentoo-r2 #1 SMP Sat Feb 11 02:18:34 CET 2012 x86_64 AMD A8-3500M APU with Radeon(tm) HD Graphics AuthenticAMD GNU/Linux
mesa 7.11.2 
xf86-video-ati 6.14.3
xorg-server 1.11.2-r2
libdrm 2.4.27

Here:

Linux amber 3.2.1-gentoo-r2 #1 SMP PREEMPT Sat Jan 28 14:05:55 CET 2012 x86_64 Intel(R) Core(TM) i5 CPU 750 @ 2.67GHz GenuineIntel GNU/Linux
(crash on kernel 3.3)
media-libs/mesa-8.0_rc1
git x11-drivers/xf86-video-ati
xorg-server 1.11.2-r2
libdrm 2.4.27

But on my radeon HD 5450, both screen is up at the kernel startup (kms), on gpu SUMO, just my first screen

Same with x11-libs/libdrm-2.4.30.

No news? Other product with be out, and card serie 7000 included.

Hi,

i have the same problem.

- Debian Sid 64bit
- Kernel 3.3.0-rc6-amd64 (Debian Experimental)
- Radeon Driver 1:6.14.3-2

Grub and kernol space boot are displayed on both displays (DVI-0 & VGA-0).
After the radeon driver is loaded VGA-0 shuts down and will not start again.
xrandr shows me both screens but only DVI-0 works.

I attach:
- dmesg
- Xorg.0.log
- xrandr output
- lspci -vv

Created attachment 58659 [details]
lspci -vv output

Created attachment 58660 [details]
Shin's dmesg output

Created attachment 58661 [details]
xrandr output

Created attachment 58662 [details]
Shin's Xorg.0.log

No body for that's?

You need this patch to fix VGA:
http://cgit.freedesktop.org/~airlied/linux/commit/?h=drm-fixes&id=700698e7c303f5095107c62a81872c2c3dad1702

I have just tested with daily buld of ubuntu and kernel 3.4.

Exactly the same in my case. No vga in my case.

No body to resolv that's?

Have you tried the patch or a 3.5 kernel?

Yes, the patch solve nothing because I don't use the VGA (HDMI + DVI, see the firt comment).

The kernel 3.5 some nothing...

(In reply to comment #19)
> Yes, the patch solve nothing because I don't use the VGA (HDMI + DVI, see the
> firt comment).
> 
> The kernel 3.5 some nothing...

Solve nothing

Can you post your dmesg and xrandr output from the 3.5 kernel?

I can before this night, because I use my computer/screen as work.

Please, I don't wish test each month on each new kernel, without any change, and at each re-ask listen: "test on the new kernel, send my the new ...".
If new patch can change my situation, please tell it, else I will provide all the test/info to solve this problem.

I wait more serious about the support. Mainly because this bug is open since 6 months, I not see any progress, and I totally blocked into my work.

(In reply to comment #22)
> I can before this night, because I use my computer/screen as work.
> 
> Please, I don't wish test each month on each new kernel, without any change,
> and at each re-ask listen: "test on the new kernel, send my the new ...".
> If new patch can change my situation, please tell it, else I will provide all
> the test/info to solve this problem.

There are a lot of changes and fixes for each new kernel.  It's not always easy to say which one(s) may help.

> 
> I wait more serious about the support. Mainly because this bug is open since 6
> months, I not see any progress, and I totally blocked into my work.

It's hard to say what your bug is.  In many cases a new kernel may help.  Dual digital connectors work fine on the sumo systems I have access to.  Also, to clarify, the DVI monitor is digital correct?  DVI-VGA adapters won't work because the DVI ports on these systems do not support analog.

Ok, to be clear:
All the display data is transmitted as numeric (no analog used). It's DVI-D + HDMI.
I connect my sumo GPU EXACTLY like I connect on my Radeon HD 5450 used actually in my computer (work with this card).

Also does manually turning the DVI-D display off/on with xrandr help?  Or boot with one display attached and plug in the second one later?  Does it work in that case?  Also what modes are you trying to use each monitor?  The xorg log you attached doesn't have much content.  If you are trying to use a high res monitor that requires duallink DVI, you will need to enable duallink support in your bios setup.  As it requires two links, enabling duallink dvi the bios also disables the secondary display on the motherboard since both are then used for DVI.

Also, are there any cases where the DVI port works?  Is it only when you try and enable both monitors that things break or does DVI not work at all under any circumstances?

I have try all of that's. In all case, I'm not able to start the second screen.
I have the same resolution with my hd 5450 or with my sumo gpu: Full HD (1920x1080).

In single screen DVI and HDMI work. Under the bios (EFI) + grub, the both screen is on.

The problem is when the both screen is connect, under linux all show the DVI screen as on, but in reality is off.

I have try lot of combinaison to start up the screen (DVI first alone, both after -> HDMI is start-up, the DVI is shutdown)

The kernel config:
CONFIG_STANDALONE=y
CONFIG_PREVENT_FIRMWARE_BUILD=y
CONFIG_FW_LOADER=y
CONFIG_FIRMWARE_IN_KERNEL=y
CONFIG_EXTRA_FIRMWARE="radeon/CEDAR_me.bin radeon/CEDAR_pfp.bin radeon/CEDAR_rlc.bin radeon/REDWOOD_me.bin radeon/REDWOOD_pfp.bin radeon/REDWOOD_rlc.bin radeon/SUMO_me.bin radeon/SUMO_pfp.bin radeon/SUMO_rlc.bin radeon/SUMO2_me.bin radeon/SUMO2_pfp.bin radeon/REDWOOD_me.bin radeon/REDWOOD_pfp.bin radeon/REDWOOD_rlc.bin"
CONFIG_EXTRA_FIRMWARE_DIR="/lib/firmware/"

Created attachment 62834 [details]
alpha_one_x86's Xorg.0.log

Created attachment 62835 [details]
alpha_one_x86's xrandr

Created attachment 62837 [details]
alpha_one_x86's kernel 3.5

I have send all the requested new file. No change...

Any idea?

It's very very hard in my case to test. Have you all to try resolv this?

@Alex Deucher: Do you work with exactly the same hardware?
CPU + GPU: A6-3650
Mother board: F1A75-I DELUXE
???

The mother board have EFI, I have no see anything into the config witch can block the graphic card.

(In reply to comment #33)
> @Alex Deucher: Do you work with exactly the same hardware?
> CPU + GPU: A6-3650
> Mother board: F1A75-I DELUXE
> ???

I don't have that exact board.

> 
> The mother board have EFI, I have no see anything into the config witch can
> block the graphic card.

Does the bios provide an option to boot via legacy bios rather than EFI?

No, but I suspect I'm in bios mode, because I use the grub2 into dos mode, and no trace of EFI in the dmesg...

I can do few donation (<20$) to help to buy the hardware to try to resolv it.
I don't know what I can do more.

Created attachment 63735 [details]
Both card

Created attachment 63736 [details]
both card enabled

I have flashed by bios, try with dual card internal + radeon HD 5450...

The graphic card have firmware? I can update it? How?

(In reply to comment #40)
> The graphic card have firmware? I can update it? How?

It's loaded by the driver.  It's used for acceleration and interrupts.  It does not affect the display hw.

Same with the kernel 3.6.2, I will donate my hardware, and change for more modern. Then I don't will test more, sorry but the problem is since lot of time (2 years?).

Created attachment 88413 [details]
dmesg with kernel 3.11

dmesg with new kernel 3.11, continues with bug...

Does selecting a different mode or refresh rate for the problematic monitor fix it?  Select something so that the displays are using different clocks.

Created attachment 88451 [details] [review]
possible fix

Does this patch fix the issue?

No, don't fix it. I can't test the patch because my try of kernel compilation have failed (Bios ROM not found). Then if you have compiled kernel to test... (without initrd, and drivers in =Y)

Or if you can provide me a live cd with this kernel patch to test.

-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/xorg/driver/xf86-video-ati/issues/26.